<br><br>2007/1/9, Thomas Friedrichsmeier <<a href="mailto:thomas.friedrichsmeier@ruhr-uni-bochum.de">thomas.friedrichsmeier@ruhr-uni-bochum.de</a>>:<br>> Hi!<br>> <br>> Since there are only two of us actively committing to SVN at the moment, and
<br>> we both have our feature changes in, I rushed the feature freeze, and started<br>> the release process for 0.4.5. More on the details of that below.<br>> <br>> First, however, I'd like to announce a preview release of
0.4.5:<br>> <a href="http://rkward.sourceforge.net/temp/rkward-0.4.5pre1.tar.gz">http://rkward.sourceforge.net/temp/rkward-0.4.5pre1.tar.gz</a><br>> <br>> Major changes in this release include:<br>> - many new plugins
<br>> - rework of the code generated by most plugins<br>> - added help system (with only some few pages, so far)<br>> - several small fixes and improvements as seen in the ChangeLog at the end of<br>> this mail.
<br>> <br>> One change in particular I'd like to get your feedback on (if you don't like<br>> it, it's not too late to remove it from the release, yet):<br>> The "Edit", "View", and "Run" menus, as well as the "Run" toolbar are now
<br>> always shown, regardless of which part/window is currently active.<br>> The "Output" menu was removed entirely ("Flush Output" was moved to Edit,<br>> and "Refresh Output" was move to View).
<br>> The idea behind these changes is, that now the menus will no longer "jump<br>> around" whenever you switch, e.g. between a help window and an editor window.<br>> The contents of the menus will still change, but the position remains fixed.
<br>> Does this seem like a good idea, or do you think it causes more harm than<br>> good?<br>> <br><br>I find it good. However I would suppose the contents of the menus to be stable.<br><br>> Of course besides this, please do some testing on the preview release, so we
<br>> can fix the most serious bugs before the official release.<br>> <br><br>Today I played a little with correlation matrix plugin and I found some different results than the ones waiting. Seeing the code I saw that there is the use of the "
pairwise.complete.obs" which means that the correlation between each pair of variables is computed using all <span style="font-weight: bold; font-style: italic;">complete</span> pairs of observations on those variables. I have a dataset where na values must not be excluded, but replaced with zero which I have done using impute of Hmisc package.
<br><br>Now to come to the suggestion. Sould the user be warned about this? Or can the user choose what to do with the na values?<br><br>I also would like to suggest for the future another method for correlation which gives the probabilities too. There are two functions doing this in two different packages in my knowledge.
<br>The one is rcorr from Hmisc and the other one is rcor.test from package ltm. However none of them has a good output using rk.print(), or rk.results().<br>